ACE: Adaptively Similarity-Preserved Representation Learning for Individual Treatment Effect Estimation

Yao, Liuyi; Li, Sheng; Li, Yaliang; Huai, Mengdi; Gao, Jing; Zhang, Aidong

doi:10.1109/icdm.2019.00186

Cited by 35 publications

(46 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With this approach, we outperform all competing algorithms on each benchmark dataset. Probably the most relevant of these comparisons are the three works [25,26,29], which generate overlapping representations of the treatment and control groups by means of neural networks and hand-designed inter-distributional similarity metric functions. Compared to them, the improvement in performance better highlights the predictive power of our representation.…”

Section: Prediction Performance Resultsmentioning

confidence: 99%

“…Existing works, such as [25,26], learn representations of individuals through fully connected neural networks, and their core spirit is to balance the distribution between different treatment groups by means of carefully designed metric functions. Meanwhile, more attention should be paid to the correlation both between covariates and between individuals to generate more discriminative representation.…”

Section: Self-supervised Transformermentioning

confidence: 99%

“…We compare CETransformer with a total of 12 algorithms. First we evaluate least squares regression using treatment as an additional input feature (OLS/LR 1 ), we consider separate least squares regressions for each treatment (OLS/LR 2 ), we evaluate balancing linear regression (BLR) [13], k-nearest neighbor (k-NN) [6], Bayesian additive regression trees (BART) [5], random forests (R-Forest) [4], causal forests (C-Forest) [23], treatment-agnostic representation network (TAR-NET), counterfactual regression with Wasserstein distance(CARW ASS) [20], local similarity preserved individual treatment effect (SITE) [25], adaptively similaritypreserved representation learning method for Causal Effect estimation (ACE) [26], deep kernel learning for individualized treatment effects(DKLITE) [29].…”

Section: Competing Algorithmsmentioning

confidence: 99%

See 2 more Smart Citations

CETransformer: Casual Effect Estimation via Transformer Based Representation Learning

Guo¹,

Zheng²,

Liu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Treatment effect estimation, which refers to the estimation of causal effects and aims to measure the strength of the causal relationship, is of great importance in many fields but is a challenging problem in practice. As present, data-driven causal effect estimation faces two main challenges, i.e., selection bias and the missing of counterfactual. To address these two issues, most of the existing approaches tend to reduce the selection bias by learning a balanced representation, and then to estimate the counterfactual through the representation. However, they heavily rely on the finely hand-crafted metric functions when learning balanced representations, which generally doesn't work well for the situations where the original distribution is complicated. In this paper, we propose a CE-Transformer model for casual effect estimation via transformer based representation learning. To learn the representation of covariates(features) robustly, a self-supervised transformer is proposed, by which the correlation between covariates can be well exploited through self-attention mechanism. In addition, an adversarial network is adopted to balance the distribution of the treated and control groups in the representation space. Experimental results on three real-world datasets demonstrate the advantages of the proposed CETransformer, compared with the state-ofthe-art treatment effect estimation methods.

show abstract

Section: Prediction Performance Resultsmentioning

confidence: 99%

Section: Self-supervised Transformermentioning

confidence: 99%

Section: Competing Algorithmsmentioning

confidence: 99%

See 1 more Smart Citation

CETransformer: Casual Effect Estimation via Transformer Based Representation Learning

Guo¹,

Zheng²,

Liu³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…By re-weighting the data samples with the estimated propensity score, we can alleviate the selection biases in observations. As for covariate balancing, a series of works solve the problem by reformulating the counterfactual estimation problem as the domain adaptation task [50,51]. Moreover, Shalit et al [42] have proved that the estimated causal effect error can be bounded by the generalization loss and the distribution distance between the treated and control groups.…”

Section: Related Work 21 Causal Effect Estimationmentioning

confidence: 99%

CausCF: Causal Collaborative Filtering for RecommendationEffect Estimation

Xie,

Liu,

et al. 2021

Preprint

View full text Add to dashboard Cite

To improve user experience and profits of corporations, modern industrial recommender systems usually aim to select the items that are most likely to be interacted with (e.g., clicks and purchases). However, they overlook the fact that users may purchase the items even without recommendations. The real effective items are the ones which can contribute to purchase probability uplift. To select these effective items, it is essential to estimate the causal effect of recommendations. Nevertheless, it is difficult to obtain the real causal effect since we can only recommend or not recommend an item to a user at one time. Furthermore, previous works usually rely on the randomized controlled trial (RCT) experiment to evaluate their performance. However, it is usually not practicable in the recommendation scenario due to its unavailable time consuming. To tackle these problems, in this paper, we propose a causal collaborative filtering (CausCF) method inspired by the widely adopted collaborative filtering (CF) technique. It is based on the idea that similar users not only have a similar taste on items, but also have similar treatment effect under recommendations. CausCF extends the classical matrix factorization to the tensor factorization with three dimensions-user, item, and treatment. Furthermore, we also employs regression discontinuity design (RDD) to evaluate the precision of the estimated causal effects from different models. With the testable assumptions, RDD analysis can provide an unbiased causal conclusion without RCT experiments. Through dedicated experiments on both the public datasets and the industrial application, we demonstrate the effectiveness of our proposed CausCF on the causal effect estimation and ranking performance improvement. CCS CONCEPTS• Information systems → Collaborative filtering; • Computing methodologies → Causal reasoning and diagnostics.

show abstract

“…For example, the shift-invariant representation learning is combined with re-weighting methods [15]. A local similarity preserved individualized treatment effect (SITE) estimation method [40,41] is proposed focusing on local similarity information that provides meaningful constraints on individual treatment estimation. Generative adversarial networks [43] for individualized treatment effects have also been proposed for individual treatment effect estimation.…”

Section: Related Workmentioning

confidence: 99%

Matching in Selective and Balanced Representation Space for Treatment Effects Estimation

Chu

Rathbun

2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

Self Cite

View full text Add to dashboard Cite

The dramatically growing availability of observational data is being witnessed in various domains of science and technology, which facilitates the study of causal inference. However, estimating treatment effects from observational data is faced with two major challenges, missing counterfactual outcomes and treatment selection bias. Matching methods are among the mostly widely used and fundamental approaches to estimating treatment effects, but existing matching methods have poor performance when facing data with high dimensional and complicated variables. We propose a feature selection representation matching (FSRM) method based on deep representation learning and matching, which maps the original covariate space into a selective, nonlinear, and balanced representation space, and then conducts matching in the learned representation space. FSRM adopts deep feature selection to minimize the influence of irrelevant variables for estimating treatment effects and incorporates a regularizer based on the Wasserstein distance to learn balanced representations. We evaluate the performance of our FSRM method on three datasets, and the results demonstrate superiority over the state-of-the-art methods. CCS CONCEPTS • Information systems → Data mining; • Computing methodologies → Causal reasoning and diagnostics; Machine learning.

show abstract

ACE: Adaptively Similarity-Preserved Representation Learning for Individual Treatment Effect Estimation

Cited by 35 publications

References 17 publications

CETransformer: Casual Effect Estimation via Transformer Based Representation Learning

CETransformer: Casual Effect Estimation via Transformer Based Representation Learning

CausCF: Causal Collaborative Filtering for RecommendationEffect Estimation

Matching in Selective and Balanced Representation Space for Treatment Effects Estimation

Contact Info

Product

Resources

About